Somatic symptom disorder

CP10CH13-Rief ARI 11 February 2014 10:45
How to Use the New DSM-5
Somatic Symptom Disorder
Diagnosis in Research and
Practice: A Critical Evaluation
and a Proposal for
Modifications
Winfried Rief1
and Alexandra Martin2
1
Department of Psychology, Philipps University of Marburg, D-35032 Marburg, Germany;
email: rief@uni-marburg.de
2
Faculty of Educational and Social Science, University of Wuppertal, D-42097 Wuppertal,
Germany; email: martin@uni-wuppertal.de
Annu. Rev. Clin. Psychol. 2014. 10:339–67
First published online as a Review in Advance on
January 2, 2014
The Annual Review of Clinical Psychology is online at
clinpsy.annualreviews.org
This article’s doi:
10.1146/annurev-clinpsy-032813-153745
Copyright c 2014 by Annual Reviews.
All rights reserved
Keywords
somatic symptom disorder, somatoform disorder, pain disorder,
hypochondriasis, illness anxiety disorder, somatization, medically
unexplained symptoms
Abstract
The fifth edition of the Diagnostic and Statistical Manual of Mental Disorders
(DSM-5) changed the term “somatoform disorders” to “somatic symptom
and related disorders” and further modified diagnostic labels and criteria.
We review evidence for the validity of the new criteria, specifically of so-
matic symptom disorder (SSD), and present a critical discussion of unsolved
and new problems. We also provide an update of mechanisms and inter-
ventions that have been empirically evaluated in somatoform disorders. For
many mechanisms, it is unclear whether their role can be easily transposed
to SSD. Therefore more research is needed on the similarities and differ-
ences between medically unexplained and medically explained conditions.
To overcome the obvious shortcomings of the current classification, we of-
fer a modification of this DSM-5 section as well as a crossover system to
apply these criteria for somatic symptom and related disorders. This pro-
posal allows working with DSM-5 but also continuing successful lines of re-
search with concepts such as hypochondriasis/illness anxiety, chronic pain,
and medically unexplained versus medically explained syndromes.
339
Annu.Rev.Clin.Psychol.2014.10:339-367.Downloadedfromwww.annualreviews.org
Accessprovidedby50.206.114.55on03/04/16.Forpersonaluseonly.
Click here for quick links to
Annual Reviews content online,
including:
• Other articles in this volume
• Top cited articles
• Top downloaded articles
• Our comprehensive search
FurtherANNUAL
REVIEWS

Contents
INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
REPLACING SOMATOFORM DISORDERS WITH SOMATIC SYMPTOM
DISORDER: THE NEWER, THE BETTER? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
The Problems with DSM-IV Somatoform Disorders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 341
How Valid Are the New Classification Criteria?. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 342
Shortcomings of the New SSD Diagnosis in DSM-5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
WHAT ARE THE MECHANISMS INVOLVED IN SOMATIC
SYMPTOM DISORDER? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 348
Cognitive-Perceptual Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
Emotion Regulation and Early Adverse Life Experiences . . . . . . . . . . . . . . . . . . . . . . . . . 353
Placebo and Nocebo Research as a Paradigm for Symptom
Development and Symptom Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 355
HOW TO USE THE NEW DIAGNOSES OF SOMATIC SYMPTOM
DISORDERS IN RESEARCH: A PROPOSAL FOR MODIFYING
DSM-5 CRITERIA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
THE NEW PROPOSAL OF SOMATIC SYMPTOM DISORDER
AND ITS RELEVANCE FOR INTERVENTION RESEARCH. . . . . . . . . . . . . . . . 359
CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
INTRODUCTION
Somatic symptoms are a common phenomenon and the most frequent reason for doctor visits.
Many somatic symptoms are neither a mere correlate of mental disorders nor medically explained
in the context of a general medical condition. Up to 80% of the general population reports so-
matic symptoms during the last seven days, and about one-fifth of the general population suffers
from serious, disabling, and frequently chronic somatic complaints (Hiller et al. 2006). Medi-
cally unexplained somatic symptoms produce health care burdens that are comparable to anxiety
and depressive disorders (Konnopka et al. 2012). Because of the tremendous relevance for health
care systems, a reliable and valid classification for somatoform symptoms is crucial as the basis
for improved detection, adequate treatment, and relevant research efforts. However, the classi-
fication for somatoform symptoms in the fourth edition of the Diagnostic and Statistical Manual
of Mental Disorders (DSM-IV) had been frequently criticized, and the fifth edition of the DSM
(DSM-5) suggests a new classification approach, summarized as “Somatic Symptom and Related
Disorders.” Although the introduction of DSM-5 was celebrated with tremendous enthusiasm
from its proponents (Kupfer et al. 2013), others expressed criticisms about missed opportunities
and uncontrolled risks (Frances & Widiger 2012). Will DSM-5 offer an improved platform for
diagnoses, management, and research in the field of somatic complaints? And what should be
done with the research results and traditions that were based on former classification approaches,
such as mechanisms of somatization, medically unexplained pain conditions, or treatment ap-
proaches for hypochondriasis? Can these results be generalized to the new diagnoses, and can
we save some successful traditional research lines using crossovers to the new classification ap-
proach? In the following article, we discuss these questions and offer specifications and crossovers
that preserve the research evidence from past classification systems while allowing the use of the
new DSM-5.
340 Rief ·Martin

SSD: somatic
symptom disorder
REPLACING SOMATOFORM DISORDERS WITH SOMATIC SYMPTOM
DISORDER: THE NEWER, THE BETTER?
The Problems with DSM-IV Somatoform Disorders
It was generally agreed by researchers that the concept of somatoform disorders introduced in
DSM-III and modified for DSM-IV needed substantial revision. Somatic symptoms that are not
attributable to biomedical conditions are a substantial problem. It is estimated that at least every
fifth visit to medical doctors is caused by this type of complaint (Steinbrecher et al. 2011). An
estimate based on data of 514 million European citizens ranks somatoform disorders third on
the list of the most prevalent mental disorders after anxiety disorders and depression (Olesen
et al. 2012). In contrast to these large prevalence rates, the prototype of somatoform disorders,
somatization disorder, is over-restrictively defined. Moreover, Anglo-American countries, mainly
the United States and the United Kingdom, rarely use the somatoform disorders diagnoses.
Statistics based on 1.2 million inhabitants of Virginia who sought medical help and received a di-
agnosis revealed that only 0.02% received a diagnosis from the category of somatoform disorders
(Levenson 2011), and a similarly low number received the diagnosis “psychological factors
affecting medical conditions.” This finding is in sharp contrast to the fact that nearly every
American family doctor reports that these patients are seen frequently in primary care (Dimsdale
et al. 2011). Many researchers were concerned that the terms somatoform and somatization
overemphasized mind-body dualism. Many clinicians interpreted these terms as indicating that
“everything is just in the mind.” Notably, DSM-III could not be blamed for this interpretation,
as it introduced the title just as a descriptive term in the sense of “somatoform = indicating a
somatic condition, albeit biomedical disorders do not explain the symptoms.”
Many physicians in Anglo-American countries found it unacceptable for their patients to re-
ceive these diagnoses (Dimsdale et al. 2011) and were concerned that patients disliked the terms,
even though physicians in other countries used these diagnoses without any patients protesting.
One deeper problem might be that guidelines of how to explain these types of diagnoses to pa-
tients were missing. Underdiagnosis is also evident for the related functional syndromes: General
practitioners did not identify 90% of patients with chronic fatigue syndrome, 77% of patients
with fibromyalgia, and 69% of patients with irritable bowel syndrome (Warren & Clauw 2012).
Patients were identified using self-reports of physicians’ diagnoses. Finally, bias in health care
systems and legal regulations can further amplify the reluctance to use a diagnosis such as so-
matoform disorder. Many health care systems tolerate underdiagnosis and misdiagnosis of mental
disorders, even though it is unacceptable to miss physical diagnoses, even those that are benign.
The DSM-5 Somatic Symptoms Disorders Work Group tried to address these issues when
developing a proposal for a new diagnostic group. They created a new category name, “Somatic
Symptom and Related Disorders,” introduced its prototype, the somatic symptom disorder (SSD),
and modified the criteria substantially (Am. Psychiatr. Assoc. 2013).
Compared to somatoform disorders and especially its prototype somatization disorder, Somatic
Symptom and Related Disorders brings substantial changes. Most radical is the abolition of the
distinction between medically explained and medically unexplained somatic complaints. As soon
as any somatic complaint qualifies as being distressing, criterion A of SSD (“One or more somatic
symptoms . . .”; Am. Psychiatr. Assoc. 2013, p. 311) is fulfilled, even if the complaint is only a
single symptom and even if it is clearly medically explained.
One critique of somatoform disorders was the lack of psychological criteria that qualified the
classification of somatic symptoms as “mental” disorders, albeit the symptoms are physical by
nature (Rief & Isaac 2007). A common characteristic of mental disorders according to DSM is
the inclusion of psychological features, e.g., body dissatisfaction in eating disorders and fear of
www.annualreviews.org • DSM-5 Somatic Symptom Disorder Diagnosis 341

fear in panic disorder. In contrast, somatoform disorders did not have any positive psychological
criteria but were based only on somatic symptoms and the “negative” criterion that they should
not be (fully) medically explained. The DSM-5 diagnosis of SSD now requires as a B criterion
that “excessive thoughts, feelings, or behaviors are related to the somatic symptoms or associated
health concerns” (Am. Psychiatr. Assoc. 2013, p. 311) and suggests that at least one of three
psychological symptoms needs to be present to fulfill the criterion. Indeed, the only justification
for classifying somatic complaints as mental disorders is the inclusion of “positive” psychological
criteria (Voigt et al. 2010).
The corresponding DSM-5 Somatic Symptoms Disorders work group obviously attempted to
limit the number of diagnoses in this category and to lump as many diagnoses as possible in one
single diagnosis (Dimsdale et al. 2013). Therefore, the new diagnosis of SSD is designed not only
to cover patients with somatization disorder but also patients with chronic pain conditions, most
patients with hypochondriasis, and many patients with medical conditions that are accompanied
by psychological features. Although the general attempt to reduce the number of diagnoses in
DSM is commended, the lumping of these heterogeneous conditions must be critically discussed.
The specificity of diagnoses should follow the same rules for all DSM diagnoses, and therefore the
question arises whether SSDs and hypochondriasis should be lumped as suggested, while major
depressive disorder, persistent depressive disorder, and other specified depressive disorder are
considered different diagnoses. In fact, the DSM-5 Somatic Symptoms Disorders Work Group
was not able to transpose this lumping process and had to acknowledge that significant groups of
patients were not covered by the diagnosis of SSD, albeit they should be. Therefore, additional
diagnoses were added (e.g., illness anxiety disorder). However, this resulted in serious problems.
If patients suffer mainly from illness anxiety, they are now disseminated over two diagnoses: If
they have an additional somatic complaint, the SSD diagnosis will be given; if they do not have an
additional somatic complaint, the diagnosis of illness anxiety disorder has to be used. This example
underlines the need to discuss these DSM-5 criteria more critically.
How Valid Are the New Classification Criteria?
The first question is whether a unique diagnosis for these syndromes is justified. Former concepts
considered somatic symptoms as a subsyndrome of other disorders (e.g., somatized depression).
This has been underlined by a substantial overlap between somatization, depression, and anxiety.
However, each of these syndromes has its own characteristics, and somatization is relevant as a
comorbid condition with other mental disorders as well as a unique syndrome (Barsky et al. 2005,
Löwe et al. 2008). Neuroimaging results and psychobiological properties further underline the
validity of a diagnosis for these somatic syndromes (Browning et al. 2011, Rief et al. 2010a). Of
note, a statistical clustering of psychopathological features resulted in a best-fitting model that
had five factors, and somatoform disorders was one of them (Kotov et al. 2011). Therefore, the
DSM-5 had good reasons to continue proposing diagnoses for this field.
The first studies evaluating SSD criteria confirmed aspects of construct and descriptive validity
as well as one-year stability of diagnoses (Rief et al. 2011b, Voigt et al. 2012, Wollburg et al. 2013).
Moreover, the initial field trials found good interrater agreement for the pilot version “complex
somatic symptom disorder” (Regier et al. 2013). However, these general aspects do not indicate
that the current proposal is the best solution. Therefore, we discuss the validity of the A and
B criteria of SSD in more detail below.
The A criterion of SSD. The A criterion allows the new diagnosis SSD to be considered as
soon as a single somatic symptom is present. The justification for this criterion is the fact that
342 Rief ·Martin

MUS: medically
unexplained symptoms
some people can suffer tremendously just from a single symptom whereas others might cope quite
well with multiple symptoms. Moreover, the distinction between medically explained symptoms
and medically unexplained symptoms (MUS) has been omitted, which seems like an adventur-
ous reformulation. The major reason for this revision is the low interrater reliability and validity
of physicians’ judgments on whether or not somatic symptoms are medically explained (Fink
et al. 2005). An analysis of MUS versus general medical symptoms did not reveal that the former
would result in higher specificity, and therefore the authors suggested using the simpler crite-
rion “physical symptoms” without necessarily identifying the cause of it (Tomenson et al. 2010,
2013). Moreover, many symptoms transform back and forth between being considered medically
explained or unexplained (Klaus et al. 2013). In many cases, the physicians’ personality rather than
the clinical picture seems to affect whether symptoms are deemed to be based on a biomedical
condition. Some physicians think of symptoms like back pain as almost always medically caused,
whereas others consider them mainly psychosomatic. Cultural influences also account for this low
interrater reliability of causality ratings.
However, the question arises whether the omission of the criterion “medically unexplained” is
the best solution for this problem. In the case examples suggested in Table 1, all three patients
would receive the diagnosis “Somatic Symptom Disorder,” yet it seems unclear whether the
mechanisms that are involved in the psychological problems are always the same and whether the
patients need the same treatments (which would be reasons to justify the use of a single diagnosis).
The lumping of the different conditions shown in Table 1 would be justified only if it had been
shown that the mechanisms were the same. However, we are not aware of studies that sufficiently
demonstrate this. Therefore, despite the newly proposed SSD, it might be useful to consider
this distinction in research and to use a modified descriptor for “medically unexplained” that was
typically used in former classification systems: The symptoms are not better explained by a general
medical condition.
Table 1 Is this the same clinical problem? Three case examples of somatic symptom disorders
Case 1: patient diagnosed with the former (DSM-IV) somatization disorder
The 43-year-old male patient reported a long history of multiple somatic complaints, starting with noncardiac chest pain and
headache when he was in school, which developed to a broad spectrum of symptoms all over the body (back pain, bloating,
abdominal discomfort, joint pains, dizziness, fatigue, general weakness). During the past 20 years, the patient received regular
medical examinations. On average, he attended a physician’s office every two weeks, and he was also admitted to hospitals at least
once or twice a year for thorough medical examinations. Apart from minor descriptive diagnoses without any pathological
significance, no biomedical explanation for the symptoms was found.
Case 2: cancer patient
The female patient developed breast cancer when she was 44. She received the typical medical treatments (surgery, radiation,
chemotherapy) and started aromatase inhibitors treatment subsequently. Now, three years later, she suffers from somatic symptoms
that could be attributed to side effects of the medical treatment (arthralgias) as well as other symptoms, such as headache. Her
major problem is that she is concerned about the recurrence risk of breast cancer. The fears of recurrence are always on her mind.
Case 3: diabetes patient
The 64-year-old man received the diagnosis of diabetes several years ago. Although he is aware that diabetes is not life threatening
as long as he adheres to the medical regime, his attention is highly focused on his body because he is unsure whether symptoms are
caused by hypoglycemia. He suffers from a multitude of somatic complaints such as joint pain, back pain, chest pain, and dizziness,
and he is concerned about developing retinopathy and finally blindness. The doctor told him that some of his symptoms could
indicate the development of diabetic neuropathy, yet symptoms and the physical conditions must be observed further before
drawing clear conclusions about the diagnosis.

Despite good reasons to use this diagnosis for patients with a single somatic complaint, we
should be aware that symptom counts still provide relevant information. First, the presence of
multiple symptoms reduces the range of medical conditions that could explain the broad variety
of somatic complaints. More importantly, the more symptoms patients report, the more likely it is
that symptoms will persist and will be associated with serious disability and reductions of quality of
life (Chou & Shekelle 2010, Rief & Rojas 2007). Vice versa, if patients report a single symptom (and
especially if it is not chronic), the likelihood is substantial that this symptom will vanish whether
medical interventions occur or not. Several studies, including latent class analyses, have confirmed
the distinction between patients with one or few symptoms and patients with many symptoms
(e.g., Rosmalen et al. 2011). Therefore, the differentiation between single- and polysymptomatic
syndromes has some validity. Even as a comorbid condition (e.g., in depression), the physical
symptom count is a significant predictor of outcome (Huijbregts et al. 2013). The same line of
research has confirmed that a differentiation into symptom clusters, such as pain syndromes,
gastrointestinal syndromes, or cardiovascular syndromes, might sometimes be found in large data
sets but does not provide any added value in classification approaches (Fink & Schröder 2010,
Rosmalen et al. 2011). Therefore, the omission of symptom clusters seems to be warranted,
although the omission of a unique diagnosis for patients with multiple somatic symptoms is less
justified.
The B criterion of SSD. The introduction of psychological and behavioral features is supported
by study results showing that it is not the somatic symptoms per se that result in increased health
care needs, but rather how patients interpret the symptoms and their illness behaviors (Rief et al.
2010b). Psychological and behavioral features are significant determinants of disability in somato-
form disorders (Wollburg et al. 2013). Although some experts suggest other ways to identify this
patient group (Fink & Schröder 2010), the inclusion of psychological and behavioral criteria makes
sense in the context of a psychiatric classification system. The B criterion of SSD requires that at
least one of three psychological criteria should be manifested.
Health anxiety. The first B criterion, health anxiety, seems reasonable, especially if the former
diagnosis of hypochondriasis is to be covered in this context. Even for many patients with somati-
zation disorder, health anxieties play a significant role (Rief et al. 2010b, 2011b), and greater health
anxiety predicts persistence of somatoform disorders (e.g., McKenzie et al. 2010). High prevalence
rates of significant health anxiety in patients seeking care in various medical specialty settings (e.g.,
neurology, respiratory medicine, gastroenterology, cardiology; mean rate 19.8% in N = 28,991)
suggests that abnormal health anxiety is a considerable problem in a broad range of somatic con-
ditions (Tyrer et al. 2011). Even after controlling for somatic symptom count, health anxiety has
an incremental effect on predicting health care use (Tomenson et al. 2012), although there are
also conflicting results (Barrett et al. 2012). Another population-based survey (German National
Health Interview and Examination Survey–Mental Health Supplement; N = 4,181) showed that
subjects with unrealistic illness worries as compared to subjects without illness worries had higher
comorbidities with medical conditions and with mental disorders, higher illness behaviors, and
lower health-related quality of life (Martin & Jacobi 2006). The impact of health anxiety in so-
matic conditions, including pain disorder, and on disability, prognosis, and treatment outcome
has been shown (e.g., Aggarwal et al. 2010, Jensen et al. 2010, Kehler & Hadjistavropoulos 2009;
for review, see Hadjistavropoulos & Hadjistavropoulos 2003). Despite the overall relevance of
health anxiety, it is not necessary for chronic somatic complaints, and many patients suffer from
somatic complaints without worrying about undetected serious medical conditions.
344 Rief ·Martin

“Disproportionate and persistent concerns about the medical seriousness of one’s symptoms.”
The second B criterion, “disproportionate and persistent concerns about the medical seriousness
of one’s symptoms” (Am. Psychiatr. Assoc. 2013, p. 311), seems to have substantial overlap with
the first B criterion, health-related anxiety, and the question arises whether this criterion brings
incremental information. Moreover, this criterion is in line with the viewpoint held by many clini-
cians but is less supported by empirical data. There are patients with a very rigid assumption about
a somatic cause of their complaints, and some of them fear serious, maybe even life-threatening,
biomedical conditions. Somatic illness attribution was found to identify those patients with
somatic symptoms who report increased disability (Rief et al. 2010b), and a few studies have shown
that the tendency for somatic causal attributions predicted disability and worse prognosis (Bailer
et al. 2008, Henningsen et al. 2005). Many overviews and textbooks describe somatizing patients as
patients who do not consider psychological causes for their symptoms. However, in reality many
patients with somatoform syndromes and/or SSD are able to consider psychological etiologies for
the complaints (for review, see Martin & Rief 2011), and the frequency of psychological symptom
attributions is very similar to somatic causality assumptions (Hiller et al. 2010). Surprisingly, some
studies have found that more psychological attributions can indicate more disabling conditions
in somatoform disorders (Rief et al. 2004). Therefore, both somatic and psychological illness
attributions can be associated with increased disability. To summarize this point, although some
evidence favors introducing this criterion, several aspects lack clarity. More precise guidelines are
needed for the application of this criterion to medically explained symptoms.
“Excessive time and energy devoted to these symptoms or health concerns.” Finally, the
third B criterion, “excessive time and energy devoted to these symptoms or health concerns,”
(Am. Psychiatr. Assoc. 2013, p. 311) has some obvious face validity, and in one of our studies,
“ruminations about bodily complaints” was found to indicate increased health care needs in
somatizing patients (Rief et al. 2010b). But again, it remains unclear whether excessive time itself
is a good criterion or whether a criterion such as “inappropriate excessive time” that is devoted to
these concerns would be more valid. If a patient has a serious risk of recurrence of life-threatening
conditions (see case example 2 in Table 1), a substantial amount of time must be devoted to
health care issues. Therefore, even if the health concerns and the time devoted to them can be
fully explained by the biomedical condition, DSM-5 suggests adding a diagnosis of SSD.
The B criterion of SSD in medically explained conditions. The introduction of “positive”
psychological criteria that define this diagnosis of SSD in DSM-5 is well supported by various
study results. It moves the focus from a mere description of symptoms to potential mechanisms.
However, whereas these criteria can be relevant and valid for MUS, it is unclear what their
role is in medically explained symptoms. Although a substantial amount of health-related anxiety
and time and energy devoted to health concerns can be considered normal or even necessary
in patients who suffer a cancer diagnosis, the same level of these features would be considered
inappropriate in patients with MUS, who may suffer from these symptoms for decades without
any new findings. Similarly, the interpretation of dizziness as a potential sign of recurrence after
suffering from a brain tumor could be well justified, whereas the same interpretation would be
considered as overcatastrophizing in case example 1 (Table 1) (somatization disorder). The DSM-
5 Somatic Symptoms Disorders Work Group decided that the suffering and potentially negative
consequences of these mechanisms are the same whether they are appropriate or inappropriate,
but this assumption definitely needs further investigation.

The specifiers of somatic symptom disorder. DSM-5 offers three specifiers for the diagnosis
of SSD that are determined by the patient’s predominant clinical syndromes. The first specifier is
“with predominant pain.” Specifiers for illness anxiety and the former somatization disorder were
under discussion in DSM-5 drafts but did not make it into the final version. When developing the
concepts of SSD, it became obvious that about 20% of patients with hypochondriasis would not
fulfill the criteria of SSD because of a lack of somatic symptoms despite serious illness anxieties.
Therefore an additional diagnosis of “illness anxiety disorder” has been introduced under the
same category.
The second specifier of SSD is “persistent,” which is defined as severe symptoms, marked
impairment, and long duration (more than six months). The duration specification is surprising
considering the fact that the C criterion already defines a similar time frame [“the state of being
symptomatic is persistent (typically more than 6 months)”] (Am. Psychiatr. Assoc. 2013, p. 311).
Finally, the third specifier allows a classification of severity, which is mainly based on the number
of psychological symptoms (B criterion). The validity of these specifiers remains unclear.
Shortcomings of the New SSD Diagnosis in DSM-5
As previously mentioned, the concept of somatoform disorders has several problems, and the
question arises whether the new formulation of SSDs is able to solve them. The first question is
whether this new concept and diagnosis will be used more frequently, especially in those coun-
tries that underdiagnose somatoform disorders (i.e., the United States and the United Kingdom).
Those experts who have criticized the risk of amplifying mind-body dualism with the old concept
of somatoform disorders might be more open to use the new diagnosis. Additionally, those who
have been reluctant to use the terms “somatoform” or “somatization” may be in favor of using
another term. However, every new term and every new concept bears the risk of being misin-
terpreted the same way the old one was, and therefore introducing a new term itself does not
solve any problems. A successful introduction heavily depends on both the people who engage
in disseminating the new diagnosis and concepts and the scientists who use it. Moreover, other
reasons for the underdiagnosis of these conditions in many patients must be addressed. The fact
that primary care doctors did not use these diagnoses can be only poorly explained by conceptual
issues and labeling. A hypothesis is that many health care providers have problems in explaining
these types of diagnoses to their patients, and more seriously, they have a greater fear that legal
issues may result from the incorrect use of this diagnosis as compared to other diagnoses. If these
reasons are relevant, then the solution is not to use a new diagnostic term, but rather to provide
training packages so that health care providers can explain these diagnoses to patients in a way
that patients can accept. Moreover, stakeholders of health care systems must be convinced that
misdiagnosing in the case of mental disorders is as serious as misdiagnosing in general.
The question arises whether the new label somatic symptom disorder is the best option for
renaming somatoform disorders (Creed et al. 2010). It is supposed to be merely descriptive (which
also was supposed to be the case for “somatoform” when it was introduced in 1980 in DSM-III). On
the other hand, this combination of terms is poorly understood by traditional psychopathologists,
and the new term still includes a substantial risk of discrimination. The psychopathological purists
mightarguethatclassificationisaprocessonthreelevels:First,thediagnosticiancollectssymptoms
(level 1), then he/she combines them to syndromes (level 2), and after the consideration of further
criteria, he/she decides whether the criteria for the disorders are fulfilled (level 3). Combining
symptoms with disorders in one term confuses the different levels of classification. Although this
can be considered to be more theoretical, the practical implications might be more serious. How
would patients feel if they had just a symptom disorder instead of having the disorder itself?
346 Rief ·Martin

How would patients feel if they had anxiety symptom disorder instead of anxiety disorders, or
depressive symptom disorders instead of depression? Some patients might feel like they do not
have a correct disorder, but just a symptom disorder. An international consensus group discussed
which terms should replace the term “somatoform” (Creed et al. 2010), and they also addressed
the international transportability of a new term; these aspects did not seem to play a role when
suggesting use of the term somatic symptom disorder.
The A criterion of SSD allows a broad variety of clinical conditions to be diagnosed in this
category, and the question arises as to whether this is not overliberal and overinclusive. The step
from the former somatization disorder to the new SSD is huge. Several studies were able to show
that up to 80% of the general population reports some somatic symptoms during the last seven
days (Hiller et al. 2006), and therefore the A criterion is easily fulfilled and can hardly be rejected
in critical cases. The well-founded validity of the distinction between single somatic symptoms
versus polysymptomatic syndromes (Lacourt et al. 2013, White 2013) also is not addressed in
DSM-5. The lack of an adequate symptom threshold in SSD was therefore considered to be a
threat to clinical utility (Voigt et al. 2010).
Moreover, if the distinction between medically explained and medically unexplained symptoms
is so difficult to draw, the omission of this criterion is only one solution. Another option could have
been to try to improve this criterion. Other diagnoses of the DSM usually use wordings such as
“not better explained by.” Without such a specification, some people fear overdiagnosing patients
with medical conditions as psychiatric cases, to subject patients to stigma, and to skew patients’
self-perception (Frances 2013b). In the same paper, Frances reported false positive rates of 7%
for SSD among healthy people of the general population, which must be considered substantial. A
criterion such as “If a diagnosed medical condition is present, the thoughts, feelings, and behaviors
are grossly in excess of what would be expected, given the nature of the medical condition” (Frances
2013a) could also be an alternative to the abolishment of the medically explained-unexplained dis-
tinction. As long as it is unclear whether the mechanisms and treatment consequences are really the
same for patients with medically explained versus medically unexplained somatic symptoms, an ad-
ditional specifier for the diagnosis indicating this distinction might be helpful, at least for research
projects. If the similarity of mechanisms and treatment consequences is confirmed, this distinction
can be eliminated, but until scientific proof is provided we should try to preserve our knowledge,
which is mainly based on MUS and not on psychological factors in medically explained syndromes.
One must also question whether the suggested psychological symptoms (the B criterion of SSD)
are the most valid, sensitive, and specific empirically founded features. They seem to fit to patients
who overlap somatization and hypochondriasis quite well, but the question arises whether the three
B criteria are highly correlated. The presented psychological symptoms are mainly cognitive-
affective. Unfortunately, behavioral features were not explicitly included in the B criteria list,
although in several fields the evidence for diagnostic validity of behavioral characteristics is higher
than for psychological characteristics. For chronic pain patients, it was shown that fear avoidance
behavior (avoidance of any situations and movements that are expected to induce symptoms) is
among the best predictors for translating acute to chronic pain (Chou & Shekelle 2010). For
patients with multiple somatic syndromes, avoiding situations that challenge the body is also one
of the most powerful variables distinguishing highly disabled patients from those with low health
care needs, even if both groups have a similar number of somatic symptoms (Rief et al. 2010b).
Behavioral avoidance is also common in some of the functional somatic syndromes, for example in
chronic fatigue syndrome, and is related to various clinical characteristics and disability (Cella et al.
2013, Nijs et al. 2013). Excessive reassurance-seeking behavior (presented to physicians as well as
to significant others, such as partners), checking behavior (e.g., self-inspection of body functions or
for signs of disease), and avoidance of situations or activities that provoke illness anxiety are major

features of hypochondriasis, and they are also often behavioral symptoms in somatization disorder.
These behavioral factors are considered crucial for the maintenance of the disorders, but none of
these features entered the symptom list of SSD B criteria. Moreover, cognitive symptoms other
than “disproportionate and persistent thoughts about the seriousness of one’s symptoms” (the first
B criterion; Am. Psychiatr. Assoc. 2013, p. 311) are not listed, whereas ruminations about physical
complaints, self-concept of bodily weakness, and subjective low symptom tolerance have previously
been suggested as cognitive variants (Rief et al. 2010b). Finally, the only affective characteristic
required for the B criterion is health anxiety; other affective symptoms such as negative affectivity,
desperation, and demoralization related to somatic symptoms are not included.
A clear critique on the DSM-5 formulation concerns splitting hypochondriasis into two
diagnoses. The clinical encounter with a patient suffering from full-blown hypochondriasis is
very different from patients with somatization disorder or patients with chronic pain conditions.
The higher relevance of anxiety processes and the more fluctuating course with ups and downs
into crises indicate that hypochondriasis has a strong overlap not only with somatization but
also with anxiety disorders. Powerful interventions have been developed for hypochondriasis; in
contrast, psychological interventions for somatization disorders and pain conditions suffer from
modest effect sizes (Clark et al. 1998, Kleinstäuber et al. 2011, Sørensen et al. 2010). These
differences between hypochondriasis and somatization underline the validity of a unique illness
anxiety diagnosis for patients with hypochondriasis. The progress in our understanding and
treatment options for hypochondriasis should not be hindered by the splitting of classification
criteria suggested by DSM-5.
The consideration of pain disorders as just a specifier of SSD must also be questioned. Reasons
in favor of this position are the significant overlap of pain symptoms with other somatic symptoms
in many patients (Fink et al. 2007, Rief et al. 2001). However, many patients report only pain
symptoms without other somatic complaints, and pain conditions are by far the most frequent
symptoms in this context. The prevalence rates for pain conditions are estimated to be above 40%
of the general population (Raspe et al. 2008). Even if combined with other somatic symptoms, pain
symptoms seem to be most stable and chronic, and they predict the most serious disability (Klaus
et al. 2013). DSM-5 reduces their tremendous societal impact to a specifier of another diagnosis.
Finally, DSM-5 also includes factitious disorder under the category of somatic symptom and
related disorders, which is definitely provoking. The initial goal of the DSM-5 Somatic Symptoms
Disorders Work Group was to reduce stigmatization and mind-body dualism of the category of
somatoform disorders, but this goal seems to be reversed when mentioning SSD in one group with
factitious disorders. We are not aware of any data showing that patients with functional syndromes
or somatoform disorders have a higher tendency to show factitious disorders. Is feigning really
more frequent in somatic syndromes than in OCD, posttraumatic stress syndrome, depression, or
panic disorder? If the inclusion of factitious disorder in this category is not supported by strong
evidence, then it is a provocation against every single patient who fears that doctors do not take
his complaints seriously, an experience that is unfortunately very common in clinical practice
(Dimsdale 2011).
WHAT ARE THE MECHANISMS INVOLVED IN SOMATIC
SYMPTOM DISORDER?
Most research on mechanisms in the SSD field has been based on the concept of somatoform
disorders, hypochondriasis, and pain disorders. The question arises whether these findings can be
transposed to the new concept of SSD. Figure 1 provides a typical example of current models
that have been developed for somatoform disorders (Kirmayer & Taillefer 1997). In the following
348 Rief ·Martin

Mechanisms
Medical illness,
physiological perturbations
Stress, trauma,
psychological disorder
Physiological disturbance Emotional arousal
Attention to body
Attribution of sensations
to illness
Illness worry,
catastrophizing,
demoralization
Communication of distress
and help-seeking
Avoidance and disability
Social response
Figure 1
A model of somatoform disorders (Kirmayer & Taillefer 1997).
sections, we introduce several suggested mechanisms and provide an update on empirical evidence
for them in reviews of recently published studies. On the basis of previous reviews and our up-
date, we discuss whether these mechanisms might also be suitable for the understanding of SSD.
(For a comprehensive overview of previous studies and detailed descriptions of the models for
somatoform disorders, see Rief & Broadbent 2007, Witthöft & Hiller 2010.) Although substantial
overlap exists between the model presented in Figure 1 and other models on somatoform disor-
ders, the other models add the role of attention systems and perceptual decisions (Brown 2004) or
attachment theory and doctor-patient interactions (Henningsen et al. 2007). In a “perceptual filter
model,” we propose processes that offer a link to psychobiological mechanisms and to bottom-up
as well as top-down regulation processes of symptom development and symptom control (Rief &
Barsky 2005). The core features of most models remain the selective attention process, the ampli-
fied perception of bodily sensations, and the perpetuation of this process through catastrophizing
interpretations and inadequate illness behavior.
Cognitive-Perceptual Processes
Perception of physical sensations. Introduced as somatosensory amplification (Barsky &
Wyshak 1990), perceptual and interpretative processes are in the center of mechanisms that
contribute to somatoform disorders. According to the concept of somatosensory amplification,

hypochondriasis in particular, but also somatization in a broader sense, results from the amplify-
ing perception of benign sensations and their misattribution to serious diseases. The tendency to
experience a somatic sensation as intense, noxious, and disturbing leads to focusing the attention
to physical signals and results again in increased awareness of bodily changes and perception of
bodily symptoms. In this way a vicious circle of symptom development and maintenance is hypoth-
esized. The empirical evidence for this assumption was mainly based on cross-sectional data and
self-rating scales (e.g., using the Somatosensory Amplification Scale; Barsky et al. 1990), showing
that the relationship between health anxiety and self-report measures of somatosensory amplifica-
tion is strong (see meta-analysis of Marcus et al. 2007). That means individuals with high degrees
of health anxiety (particularly individuals with hypochondriasis) believe that they are especially
aware of and sensitive to their bodily processes. However, it remains unclear whether they are in
fact more sensitive and accurate interoceptors. On the basis of their systematic review, Marcus
et al. (2007) concluded that there is little evidence of a positive relationship between health anxiety
and the ability to perceive autonomic processes or tactile stimuli accurately. Comparably, findings
are inconsistent regarding the perception of body signals in somatoform disorders. Although an
earlier study demonstrated more precise perceptions of muscle tension (Scholz et al. 2001), two
recent studies were unable to show more accurate interoceptive awareness in somatoform dis-
orders and in noncardiac chest pain using two well-established heartbeat perception paradigms
(Schaefer et al. 2012; S. Schroeder, A. Gerlach, S. Achenbach, & A. Martin, manuscript submit-
ted). However, both studies revealed that the severity of the disorder may be associated with lower
(rather than higher) interoceptive awareness. These findings are in line with two recent models of
somatoform disorder that stress the importance of top-down factors influencing the perception of
somatic sensations (Rief & Barsky 2005) and more specifically of biased schema-guided process-
ing of interoceptive information (Brown 2004). In recent years, a series of novel experiments has
provided accumulating evidence that the experience of MUS is indeed characterized by distorted
perceptual processing of (aversive) somatic sensations. Some of these experiments varied sensory
input and/or used perceptual illusion techniques such as the somatic signal detection task or the
rubber hand illusion (Bogaerts et al. 2010b, 2012; Brown et al. 2010, 2012; Katzer et al. 2011,
2012; Miles et al. 2011).
Bogaerts and collegues (2010b) assessed the experience of dyspnea in relation to respiratory
changes due to a laboratory-induced dyspnea in MUS. The results showed that perceived symp-
toms were less strongly related to relevant physiological parameters in MUS patients than in
healthy persons, specifically when afferent physiological input was relatively weak. This suggests
a stronger role for top-down psychological processes in symptom perception in patients with
MUS. Another experiment (Bogaerts et al. 2012) applied the recency effect paradigm, in which
unpleasant somatic experiences are remembered as less aversive when they end gradually rather
than abruptly (i.e., the peak-end effect). The authors manipulated the end of laboratory-induced
dyspnea in patients with medically unexplained dyspnea. Patients with medically unexplained dys-
pnea did not show this peak-end effect but rather a deficit recovery of dyspnea that could not
be explained by altered respiratory physiology. This experiment very elegantly demonstrates the
critical role of distorted perceptual-cognitive processing of aversive somatic sensations in MUS.
Other studies adapted the somatic signal detection task (SSDT) as an experimental paradigm
to assess touch illusions as a laboratory analog of MUS. During the SSDT, participants indicate
whether they feel vibrations delivered to their fingertip at perceptual threshold. In the experi-
ment, the presence of the vibration and of an additional light signal is varied (Lloyd et al. 2008).
Some participants report the presence of the vibration even when no tactile stimulation has been
delivered (so-called false alarms or illusory touch experiences), a response pattern that was as-
sociated with somatoform dissociation (Brown et al. 2010) and with physical symptom reporting
350 Rief ·Martin

(Brown et al. 2012). The latter study even demonstrated that no difference existed in the false alarm
rate between patients with medically explained and medically unexplained symptoms, suggesting
that the tendency to experience somatosensory distortion is linked to physical symptom reporting
(Brown et al. 2012) and therefore may also be relevant in SSD. Another research group did not
show a higher false alarm rate in patients with somatoform disorder compared to healthy controls,
but it did identify a more liberal response bias (in the first half of the light-absent condition of
the SSDT) in patients with somatoform disorder (Katzer et al. 2012). Within the somatoform
disorder group, the report of somatoform (especially pseudoneurological) symptoms correlated
positively with illusory tactile perceptions in the SSDT.
Overperception of symptoms seems to be strongly influenced by top-down processes. This
effect can be linked to emotional cues that are able to activate different symptom-related schemas
(representations) and has been investigated by comparing high- versus low-habitual symptom
reporters (Bogaerts et al. 2010a, Constantinou et al. 2013). Although all participants reported
higher levels of negative affect during the negative and symptom-related picture series compared
to the positive and neutral picture series, only high-habitual symptom reporters reported more
bodily symptoms after viewing the negative and symptom-related pictures (Bogaerts et al. 2010a).
A learned association between negative emotional states and symptom reporting in high-habitual
symptom reporters could explain this effect.
Symptom perception also seems to be strongly influenced by contextual variables. This has
been demonstrated by Pennebaker (1982), who showed that distraction reduces the intensity
of somatic symptom perception. The competition of cues model postulates that internal and
external cues (internal sensations and symptoms, sensory perception of external stimuli) compete
for the attention of the individual, and the larger the power of one of these cues, the lower the
perceptual intensity of the others. Neuroimaging techniques further confirmed that distraction
not only influences the cognitive interpretation of sensations but is also directly related to reduced
amplitudes of sensory stimulation even in primary somatosensory fields (Bantick et al. 2002).
Therefore, even in the era of acceptance strategies, distraction plays a role as a potential symptom
coping strategy (Kohl et al. 2013).
Because alterations in the perception of body signals are considered essential for the develop-
ment and maintenance of somatoform disorders, it is important to continue research regarding
perceptual accuracy versus perceptual distortions. Some of the recent findings indicate that these
processes are related to general symptom reporting independent of their etiology; therefore, these
mechanisms may also be important in SSD, especially with polysymptomatic manifestations.
Symptom interpretation and catastrophizing. Catastrophizing cognitions refer to the ten-
dency to overinterpret bodily sensations as signs of a serious disease and/or the tendency to
overinterpret the likelihood and/or intensity of potential negative consequences of symptoms. In
pain research, catastrophizing often refers more generally to the tendency to excessively focus on
pain sensation (rumination), exaggerate its threat (magnification), and to feel helpless about pain
(helplessness). Other examples of dysfunctional cognitions about illness-related constructs are the
rigid assumption that being healthy is defined by the total absence of any physical sensations,
beliefs that one has no control over symptom course or recurrence, or self-concepts of being
bodily weak and intolerant of stress. Evidence is robust for these types of dysfunctional body- and
health-related beliefs in health-anxious individuals (Marcus et al. 2007), and it also exists in other
somatoform disorders and in chronic pain conditions (e.g., Rief & Broadbent 2007). A recent
study also provided evidence that in patients with somatoform disorder the explicit self-concept is
related to bodily weakness, and the implicit health-related self-concept differs from that of healthy
control subjects (Riebel et al. 2013).

The role of cognitive interpretation of perceived sensations has been well confirmed using
correlational studies and predictor studies. Catastrophizing somatic perceptions increases the
aversive quality of the stimuli and is related to higher pain intensity and higher degrees of disability
in a variety of situations. Catastrophizing is negatively associated with the tolerance of pain and
other symptoms (e.g., Leeuw et al. 2007, Quartana et al. 2009), pain thresholds (Geisser et al. 2003),
and even correlates with brain activation caused by painful stimuli in somatosensory fields of the
cortex (Seminovicz & Davis 2006). Pain catastrophizing seems to mediate the relationship between
sensation and suffering through unpleasantness, as well as the relationship between pain-related
unpleasantness and suffering (Wade et al. 2011). Whereas many studies stress the significance of
catastrophizing cognitions in chronic pain and other somatic conditions, other research has shown
that preoperative catastrophizing predicts acute, postsurgical pain independent of the presence of
anxiety or depression (e.g., Kahn et al. 2012). An up-to-date review also reports empirical support
for an association between high catastrophizing and high fatigue in various clinical groups, e.g.,
patients with cancer, multiple sclerosis, and fibromyalgia (but because the number of studies is
small, one cannot yet draw any conclusions for the separate conditions) (Lukkahatai & Saligan
2013). Fatigue catastrophizing is one cognitive factor that possibly contributes to persistent fatigue
after cancer treatment (Goedendorp et al. 2013).
A catastrophizing interpretation bias could increase anxiety, which is associated with further
physiological changes as well as body-related hypervigilance; both of these reactions promote
further symptom perceptions. These cognitive processes could play a causal role in medically
unexplained somatic conditions. In SSD related to a known biomedical condition, these cognitive
biases may also contribute to increased symptom experiences, maladaptive illness behavior, and
disability (a role that needs to be evidenced), but possibly would be of less importance regarding
onset of the disorder itself.
Another way to assess the impact of automatic information-processing bias is to experimentally
apply the affect misattribution procedure (AMP) by manipulating qualities of prime stimuli. One
recent study provided preliminary evidence in a student sample showing that ambiguous stimuli
(i.e., Chinese characters) were less often rated as pleasant after presentation of health-threatening
prime pictures compared to conditions featuring neutral or no primes and that this AMP effect
was associated with a self-report measure of health anxiety ( Jasper & Witthöft 2013). However,
another recently completed experimental study (S. Schroeder, A. Gerlach, & A. Martin, manuscript
submitted) on automatic affective evaluations of somatosensory stimuli (modified AMP) was unable
to show a stronger automatic negative interpretation bias concerning somatosensory sensations
in patients with noncardiac chest pain when compared to subjects either with cardiac chest pain
or without chest pain. Stronger anxiety sensitivity was the only psychological factor related to
automatic affective interpretations of somatic sensations. This finding is further confirmed by a
recent longitudinal study (Ruchkin & Schwab-Stone 2013).
To summarize, further studies are required to investigate the relevance of automatic interpreta-
tive processes to the course of distressing somatic syndromes with or without clinically significant
health anxiety. Recent, more experimental approaches extend the knowledge based on correla-
tional designs but offer only initial insights into the complex mechanisms involved.
Hypervigilance to somatic sensations and attention to illness-related stimuli. Hypervigi-
lance toward bodily sensations and attentional bias for illness-related cues are central mechanisms
in hypochondriasis and other somatoform disorders, according to cognitive-perceptual models.
Attentional biases toward threat-relevant stimuli would be expected in health-anxious individuals
and may also be present in those with DSM-5 illness anxiety disorder and SSD with health anxiety.
A recent study on attentional bias for disorder-related words (emotional Stroop) comparing
352 Rief ·Martin

CBT: cognitive
behavioral therapy
patients with hypochondriasis, patients with other somatoform disorders, patients with panic
disorder, and controls showed the strongest bias for threat-related words in patients with
hypochondriasis. Results also suggest an attentional bias, although less strong, for illness-related
words in other somatoform disorders (Gropalis et al. 2012). The attentional bias was reduced after
four months of CBT for hypochondriasis and somatoform disorder but not for panic disorder.
Memory bias is another cognitive bias known to be relevant in other emotional disorders.
Some findings suggest that implicit and explicit memory biases occur in patients with somatoform
disorder (e.g., Lim & Kim 2005, Martin et al. 2007) and in those with chronic pain (e.g., Pincus
& Morley 2001). However, two recent studies did not provide additional support for a disorder-
specific memory bias in hypochondriasis and other somatoform disorders based on recognition
(Gropalis et al. 2012), free recall, or directed forgetting of illness-related word stimuli (Wingenfeld
et al. 2013). Nikendei and colleagues (2009) also failed to find a selective processing of word
stimuli related to different symptom attributions when comparing pain patients with a somatic
causal attribution style to those with a psychosocial causal attribution style. In a former study,
we demonstrated that memory bias is based less on the verbal content per se and more on the
memorized likelihood of threatening explanations (Rief et al. 2006b). The direct comparison
of subjects with similar somatic symptoms resulting from different diseases produced further
interesting results. Evidence for hypervigilance toward gastrointestinal stimuli (word recognition
and recall) was stronger in patients with irritable bowel syndrome in comparison to those with an
organic gastrointestinal disease (Posserud et al. 2009).
Evidence for cognitive mechanisms comes from investigating their variations in treatment con-
texts. One of the first evaluation studies using the SSD concept found that reductions in health
anxiety, self-concept of bodily weakness, and body scanning significantly predicted improvement
of physical functioning (Voigt et al. 2012). Results did not differ for patients with somatoform
disorders and patients with SSD, and therefore these cognitive mechanisms seem to play a role
in both concepts. However, the samples were highly overlapping and were selected from a psy-
chosomatic hospital, and it remains unclear whether a sample from a more medical setting would
produce the same results.
Emotion Regulation and Early Adverse Life Experiences
Emotion and emotion regulation have been thought to be related to somatic symptoms since
the introduction of the term “somatization” by Stekel in 1924 (Allen & Woolfolk 2013). The
alexithymia concept was developed in the 1970s and hypothesized that people who are unable to
detect, name, and express feelings have an increased risk to develop somatic symptoms. The typical
assessment instrument for alexithymia, the Toronto Alexithymia Scale (TAS or TAS-R; Taylor
et al. 1992), has shown scores to be significantly associated with the number and intensity of somatic
symptoms (Bach & Bach 1995). However, these associations between TAS scores and somatoform
symptoms could be influenced by other common factors, such as depression, negative affectivity,
or traumatic life experiences (Rief et al. 1996). Negative affectivity and neuroticism have been
considered as precursors of somatic complaints for decades (Watson & Pennebaker 1989).
Many patients with multiple somatic symptoms do not associate their symptoms with their
emotions, and a lack of emotional awareness is associated with difficulties in the differentiation of
physical symptoms (Burton et al. 2009, Subic-Wrana et al. 2010). The psychodynamic approach
postulates emotion regulation deficits in somatization (Henningsen et al. 2007); the deficits are
thought to be caused by early life experience. Modern concepts of pain management emphasize
the close association of emotion management and pain control and the symptom-enforcing role of
negative emotions (Schweinhardt et al. 2008). A recent investigation indicates that the reduction

of negative emotion predicts reduced pain, especially after a period of high negative emotion
(Connelly et al. 2007). Social exclusion can be another powerful trigger for negative emotions, and
depending on the intensity and time perspective, the effect of social exclusion on affect and physical
sensitivity versus numbing can be different (Bernstein & Claypool 2012). Emotion regulation
capacities play a unique role in the prediction of specific facets of adjustment among people with
chronic pain (Agar-Wilson & Jackson 2012). However, our understanding of the role of emotion
regulation as a risk factor, as a correlate, or as a consequence of somatic symptoms is far from being
clear, and more longitudinal and experimental studies are needed in this field. Moreover, whether
our limited knowledge on the mechanisms of emotion regulation in somatoform disorders also
applies to the new concept of SSD is the subject of vague speculations.
Early lifetime adversities are significantly higher in patients with somatoform disorders, even
when assessed using longitudinal approaches (Creed et al. 2012, Gulec et al. 2013, Kuwert et al.
2012, Nelson et al. 2012). Interestingly, an association between traumatic life events and somatic
symptoms has been found not only for somatoform disorders and hypochondriasis (Barsky et al.
1994) but also for functional somatic syndromes such as fibromyalgia (Bohn et al. 2013) and irrita-
ble bowel syndrome (Bradford et al. 2012). The role of traumatic experiences has been emphasized
in typical biomedical conditions such as breast cancer and osteoarthritis (Stein & Barrett-Connor
2000). The current psychobiological model postulates that early lifetime adversities cause ab-
normalities of the biological stress response (hypothalamic-pituitary-adrenal axis) (Heim et al.
2008a,b) that can last throughout life and lead to vulnerability of the individual to develop somatic
symptoms or disorders even decades after the original adverse events (Kuwert et al. 2012).
The experience of traumatic events also seems to cause the development of alexithymic patterns,
which is an interesting extension of the traditional alexithymia model. Hereby alexithymia is not
considered as a cause of somatic symptoms but rather as a coping strategy after experiencing over-
whelming emotional and painful situations. It could be helpful for some patients to downregulate
emotional processes after experiencing traumatic events. Increased rates of traumatic experience
are considered to be a potential cause of emotion regulation deficits (Cloitre et al. 2005). More
research is needed to understand when emotional downregulation is a helpful strategy and when
it is associated with emotional avoidance, which functions as a mechanism to prolong symptoms.
Attachment styles have also been connected to the development of somatic symptoms, especially
after the experience of adverse early life events. In particular, insecure attachment styles are
associated with an increased risk of developing somatic symptoms (Taylor et al. 2000, 2012): a
prevalence rate of 28% is reported for insecure attachment in patients with MUS. However, the
sensitivity of this criterion is obviously low (with 72% of target patients not showing insecure
attachment styles), and the specificity of such a potential risk factor is questionable. Insecure
attachment styles have been reported for many other diseases as well (McWilliams & Bailey
2010). The generalization of results on attachment styles from somatoform disorders to SSD is
currently not warranted because for many medical conditions, such an association is not shown.
The investigation of all the mechanisms reported above suffers from some serious limitations.
Mechanisms that are shown to be able to increase the likelihood that a patient will report chronic
somatic symptoms are not identical to the “real” causes for the development of these symptoms in
a specific patient. Some of the mechanisms are more relevant for the maintenance and persistence
of symptoms than for the original development of them. Some mechanisms are more relevant for
understanding symptoms, whereas others are more relevant for understanding associated distress,
disability, or lack of coping skills. Despite this critique, our understanding of mechanisms that
are involved in the field of somatoform symptoms has greatly increased over recent decades (Rief
& Broadbent 2007). The relevance for the new category of SSD is obvious for some of these
mechanisms, and the application for the new diagnosis seems justified; in contrast, several other
354 Rief ·Martin

mechanisms seem to be very specific for somatoform disorders, and their relevance for the new
diagnosis is speculative. Notably, the role of the presented mechanisms could differ in SSD in
cases with a clear biomedical disease and cases without a sufficient medical explanation. As a
potential causal factor in MUS, the same psychological factor could work as a mediator of distress
in medically explained symptoms or interact in a different way with reinforcement processes
and other mechanisms during the maintenance period. The similarities and differences of these
mechanisms in variants of SSD are not well understood, and a premature merging of medically
explained and unexplained symptoms would prevent their potential detection.
Placebo and Nocebo Research as a Paradigm for Symptom
Development and Symptom Control
Placebo and nocebo mechanism investigations are also shedding light on SSDs. While placebo
mechanisms can increase the expected benefit of a treatment, the psychological factors that re-
duce the expected treatment effects or that induce additional side effects are so-called nocebo
mechanisms. Over the past 20 years, it has been impressively shown that placebo and nocebo
mechanisms not only influence psychological aspects of disorders but also are able to influence
physiological aspects of most biological systems (Enck et al. 2013, Rief et al. 2011a). This line
of research is of relevance for SSDs because nocebo mechanisms can be understood as psycho-
logical processes in the development of somatic symptoms, whereas placebo mechanisms can be
interpreted as mechanisms to control symptom perception. The most frequent mechanisms ex-
amined in this field are the roles of expectations, associative learning, and interactional aspects
such as doctor-patient relationship. If people expect to develop symptoms after stimulations (e.g.,
through medication, electromagnetic fields, wind turbines), this expectation increases the risk
of developing these symptoms even if no further stimulations are provided. These results have
been demonstrated experimentally in a variety of studies (Crichton et al. 2013, Oftedal et al. 2007,
Vernia et al. 2013, Witthöft & Rubin 2013). Moreover, people can also learn to develop symptoms
in a specific environment through processes of associative learning (Colloca et al. 2008, 2010).
Even observational learning can play a role in amplifying the development of somatic symptoms:
the more pronounced a pain reaction of an observed person in a specific situation is, the higher
the pain perception of the observer when being confronted with the same situation; similarly, the
observation of a placebo effect increases subsequent placebo reactions of the observer (Colloca
& Benedetti 2009). The development of side effects in clinical trials seems to be another phe-
nomenon that must be interpreted under this perspective. Many patients discontinue drug intake
explicitly because of side effects that are attributed to the medication, even though these patients
are in the placebo group (Rief et al. 2006a, 2009). The development of side effects seems to be
frequently associated with personality factors, sometimes even more than with features of the
medication itself. In other words, some people seem to have learned to respond with side effects
to any medication that is presented, and they develop an expectation that they will continue to
experience side effects in the future (“side effect sensitivity”); therefore, the assessment of past side
effect experience is one of the best predictors for the development of future side effects in medical
interventions (Horne et al. 2013).
The results of placebo and nocebo research indicate that the corresponding mechanisms
are able to act in psychological as well as physiological systems. Therefore, translating these
mechanisms from somatoform disorders to SSDs seems to be justified because many studies in
the placebo field dealt with somatic symptoms in general (not only somatoform). However, the
role of expectation and associative learning can be different for every biological system as well
as for the psychological system. Pain perceptions seem to be vulnerable both to expectation

and conditioning. Physiological systems with less perceptual aspects (e.g., immune functioning)
seem to be more prone to modifications via learning principles, whereas suggestions/cognitive
instructions are less relevant in these conditions.
Placebo and nocebo mechanisms are also relevant to intervention planning. The modification
of expectations as well as extinction learning should play a significant role when patients cope with
symptoms. Unfortunately, only a few intervention programs for patients with somatoform symp-
toms focus on these mechanisms. A reformulation of intervention programs, from the perspective
of (a) how they address expectation optimization and (b) how to modify learning processes that
were acquired over the many years of symptom persistence, is necessary.
HOW TO USE THE NEW DIAGNOSES OF SOMATIC SYMPTOM
DISORDERS IN RESEARCH: A PROPOSAL FOR MODIFYING
DSM-5 CRITERIA
The new criteria for SSDs pose several problems in terms of the continuation of research fields and
transformation of current knowledge to the new diagnoses. Some diagnoses such as SSD include
an extreme heterogeneous cluster of patients with MUS, patients with medically explained symp-
toms with significant psychological factors, patients suffering from pure pain symptoms, or patients
suffering primarily from hypochondriasis. Patients with unique and specific syndromes (e.g., low
back pain) are summarized together with patients with a broad variety of multiple symptoms (e.g.,
somatization disorder). Patients with the diagnosis of SSD can belong to either a group with a
very good response to psychological interventions (e.g., hypochondriasis subtype) or a group with
low-to-moderate effect sizes of psychological interventions (e.g., chronic pain syndromes). For
some of the conditions, anxiety is a prominent psychological feature (e.g., hypochondriasis, fear
of progression in cancer), whereas for others, demoralization or physical deconditioning might be
more relevant. Moreover, the need to include medical aspects in psychological intervention pro-
grams can differ significantly for patients with the same SSD diagnosis. All these aspects challenge
the reliability and validity of the diagnostic process and raise questions about the generalizability
of research results. Finally, the question arises whether the SSD diagnosis fulfills the most basic
requirement of diagnoses, namely to facilitate communication in the health care system. The diag-
nosis of SSD, based on the current concept of the term, would not be sufficient to steer treatment
planning; health care providers would require much additional information. It is impossible to
define treatment recommendations for SSD without defining subtypes; thus, attempts to develop
empirically founded treatment guidelines would be disabled.
Until the questions raised above are answered, we suggest that the new proposals on how to
use the DSM-5 somatic symptom and associated disorders diagnoses be modified. The following
proposal enables successful fields of research and treatment recommendations to continue as well
as cross-reference to the DSM-5.
At least for research purposes, it would be useful to distinguish the following groups and to
define common classification criteria:
somatic symptom disorder, somatic symptoms not better explained by a biomedical condi-
tion, monosymptomatic type
somatic symptom disorder, somatic symptoms not better explained by a well-known biomed-
ical condition, polysymptomatic type
somatic symptom disorder in the context of well-known biomedical conditions
pain disorder (specify: monosymptomatic versus polysymptomatic type; episodic versus
chronic type)
illness anxiety disorder (hypochondriasis)
356 Rief ·Martin

Table 2 offers an overview of suggested criteria for these five clinical problems as well as a cross-
reference to the existing DSM-5 system. The validity for this suggested subclassification comes
from different treatment responses (e.g., illness anxiety versus chronic somatization), different
treatment needs (medically explained somatic symptoms versus medically unexplained somatic
symptoms), different time course of the disorder (e.g., more fluctuation in illness anxiety versus
chronic stability in polysymptomatic somatization), different prognoses (poorer prognosis for
polysymptomatic versus monosymptomatic disorders), and the different roles of behavioral and
psychological factors.
The proposal shown in Table 2 offers a valid classification approach that allows the changeover
to DSM-5 but also makes use of current research knowledge. At present, it is unclear whether SSD
in the context of a serious medical condition such as cancer or diabetes can be merged with SSD
without any biomedical pathology, and as long as this is unclear, we recommend differentiating it
in research. Notably, other diagnostic categories of DSM-5 allow for diagnoses “due to another
medical condition” (e.g., depressive disorder; Am. Psychiatr. Assoc. 2013, p. 180) or use criteria
such as “not attributable to a concurrent medical condition” (e.g., food intake disorder; Am.
Psychiatr. Assoc. 2013, p. 334); the same rationale could apply for SSD. Moreover, the prominent
psychological features in different medical conditions can also vary, and a different treatment
may be needed for a patient suffering from the fear of suffering from hypoglycemia or long-term
consequences in diabetes than for a patient suffering from cancer.
The distinction between monosymptomatic and polysymptomatic syndromes can be con-
sidered important because symptom counts and number of pain sites are significant predictors
of persistence and disability as well as work absenteeism (Escobar et al. 2010, Haukka et al.
2013, Rosmalen et al. 2011). Moreover, the number of functional somatic syndromes has been
evidenced as a strong risk factor for the incidence of another functional somatic syndrome
(Warren et al. 2013). If polysymptomatic syndromes are established, the likelihood of the
remission of symptoms and a return to normal quality of life will be very low (Keijsers et al. 2010,
Rief & Rojas 2007).
Illness anxiety disorder clearly differs from syndromes that were formerly characterized as
somatizationdisorderorchronicpaindisorders.Inmanycases,thecourseofillnessanxietydisorder
resembles the course of fluctuating anxiety disorders such as panic disorder, with hours or days of
emotional crises being followed by periods of low anxiety. Treatment response to hypochondriasis
is high, whereas treatment response to other SSDs can be low to medium. All of these aspects
underline the validity of a unique illness anxiety disorder/hypochondriasis diagnosis.
Finally, pain syndromes are by far the most prevalent single somatic symptoms, with special
relevance for chronicity and health care use, which underlines the need to establish a distinct pain
diagnosis and not just a specifier. Moreover, pain research has its own traditions and concepts,
with unique diagnostic approaches (e.g., chronic widespread pain, neuropathic pain), investigation
of outcome predictors (e.g., fear avoidance), and treatments (e.g., multidisciplinary treatments,
sometimes combining analgesic and psychological interventions). Mechanisms such as the percep-
tual sensitization process, supported by neural plasticity and activation processes in the so-called
pain matrix of the brain, are clearly defined for pain conditions but less so for the variety of (other)
syndromes of SSD (Davis & Moayedi 2013).
As mentioned above, the empirical evidence for the selection of the “positive” psychological
criteria introduced by DSM-5 is still weak. Therefore we should not wait until the dawn of DSM-6
before evaluating the classification validity of the present, and other, psychological criteria. This
has two implications: First, the suggested DSM-5 B criterion for SSD should be continuously
evaluated for its validity; and second, new criteria that could be alternatively or additionally used
should be investigated (see below).

Table 2 Proposed ways to use the DSM-5 diagnoses “somatic symptom and related disorders” in
research
Somatic symptom disorder, type A (MUS; monosymptomatic type)
Description: somatic symptoms are not better explained by a general medical condition;
monosymptomatic
How to classify?
DSM-5 criteria for somatic symptom disorder apply
One somatic symptom or well-circumscribed somatic syndrome is the focus of somatic distress (e.g.,
nausea, abdominal pain and bloating, noncardiac chest pain, dizziness)
The most significant somatic symptoms causing distress for the patient are not better explained by a
well-known biomedical condition
Specify:
Gastrointestinal type (e.g., irritable bowel syndrome, nonulcer dyspepsia)
Cardiac type (e.g., noncardiac chest pain)
Conversion type (e.g., pseudoseizures)
Fatigue type (e.g., chronic fatigue syndrome)
Environment attribution type (e.g., multiple chemical sensitivities)
Somatic symptom disorder, type B (MUS; polysymptomatic type)
Description: somatic symptoms are not better explained by a general medical condition; polysymptomatic
How to classify?
Somatic symptoms that are the focus of somatic distress cover more than one body site/biological
system (e.g., former “somatization disorder” and other patients with multiple somatoform symptoms)
The most significant somatic symptoms causing distress for the patients are not better explained by a
well-known biomedical condition
Somatic symptom disorder, type C (with biomedical condition)
Description: somatic symptoms in the context of a well-known biomedical condition
How to classify?
Somatic symptoms that are the focus of somatic distress are mainly caused by a well-known
biomedical condition
Specify: biomedical condition (e.g., diabetes, cancer, Crohn’s disease, myocardial infarction,
neurodermatitis, epilepsy, migraine)
Pain disorder
Description: Pain disorder
How to classify?
Pain symptoms are the focus of somatic distress
DSM-5 criteria for somatic symptom disorder, specifier “with predominant pain,” apply
If pain symptoms are in the context of further somatic complaints, a diagnosis of somatic symptom
disorder type B might be more appropriate
Specify:
Pain mainly caused by a well-known biomedical condition
Pain characterized by physical and psychological factors
Monosymptomatic versus polysymptomatic pain conditions
(Continued )
358 Rief ·Martin

Table 2 (Continued )
Illness anxiety disorder
Description: illness anxiety, hypochondriasis
How to classify?
Preoccupation with having or acquiring a serious illness. If somatic complaints are present, or if
illness anxieties or preoccupation are clearly excessive or disproportionate and are a significant
reason for distress
One of the following DSM-5 criteria applies:
Criteria for somatic symptom disorder
Illness anxiety disorder
Note: Although somatic symptom disorder types A and B are exclusive, type C can be combined with both of the other two
types. Pain disorder could be combined with the other somatic symptom disorder diagnoses if pain disorder itself does not
fully cover the reported symptomatology but is prominent enough to need its own diagnosis. Illness anxiety disorder can be
combined with any other disorder reported here. Abbreviations: DSM-5, Diagnostic and Statistical Manual of Mental
Disorders, fifth edition; MUS, medically unexplained symptom.
Although cognitive, affective, and behavioral characteristics are often interrelated, they are not
necessarily present at the same time and within the same individual. Rather, these features seem
to vary considerably between affected subjects. We therefore suggest further operationalizing the
B criterion of SSD by providing a range of cognitive, affective, and behavioral characteristics for
further evaluation of validity aspects.
Cognitive symptoms: disproportionate and persistent thoughts about the seriousness of
one’s symptoms (e.g., somatic causal symptom attribution despite contradicting medical
information), catastrophizing of bodily sensations, self-concept of bodily weakness, beliefs
about low symptom tolerance, fear avoidance beliefs, rumination about physical complaints.
Affective symptoms: health anxiety, fear of movement, demoralization because of somatic
symptoms, negative affectivity.
Behavioral symptoms: fear avoidance behavior, avoidance of physical effort, avoidance of
situations or activities that provoke illness anxiety, excessive reassurance-seeking behavior
(presented to physicians as well as to significant others, such as partners), checking behavior
(e.g., self-inspection of body functions or for signs of disease).
THE NEW PROPOSAL OF SOMATIC SYMPTOM DISORDER
AND ITS RELEVANCE FOR INTERVENTION RESEARCH
The DSM-5 proposal for somatic symptom and related disorders does not seem to offer a good
starting point for intervention research, whereas the modifications suggested in this review offer a
better platform for treatment studies. From a scientific point of view, treatments are most successful
if they follow the results of mechanism research. A good example is hypochondriasis/illness anxiety
disorder. Some psychological interventions follow the conceptualization that hypochondriasis is
mainly driven by anxiety and therefore interventions focus on managing anxieties: These interven-
tions have been shown to be highly successful, with effect sizes of Cohen’s d above 2.0 (Salkovskis
1995). Even short (six-hour) interventions succeed in reducing the syndrome significantly (Barsky
& Ahern 2004), and Internet-based applications confirmed very large effect sizes (e.g., d = 2.09
for health anxiety inventory, pre-treatment versus follow-up scores) (Hedman 2011).
Several approaches have been tested for various pain disorders but with only moderate effect
sizes so far (Glombiewski et al. 2010a,b; Morley 2011). Two strategies can be helpful to further

improve treatment efficiency: First, we can try to better understand the mechanisms involved in
chronic pain development and address them more specifically; second, we can try to tailor treat-
ments more specifically to individual problem profiles. Psychological pain management programs
have been shown to be useful, but they offer a broad and less theory-driven approach. A substantial
subgroup of patients with chronic pain shows fear avoidance (Vlaeyen & Linton 2000), which is
a major predictor for symptom persistence. For those patients, exposure-based interventions may
be successful (de Jong et al. 2012). Moreover, third-wave interventions such as acceptance and
commitment therapy or mindfulness-based cognitive therapy have been also shown to be effec-
tive, but on a global level they do not seem to be more effective than standard CBT procedures
(Wetherell et al. 2011). An increase in success rates might be achieved by developing decision
rules to identify which patient needs standard CBT, exposure-based interventions, and/or more
acceptance-based treatments.
For somatization disorder and multiple somatoform symptoms, the effect sizes of interventions
are low yet still significant (Kleinstäuber et al. 2011, Sattel et al. 2012, Schroder et al. 2012). The
question arises whether current treatment approaches adequately incorporate the knowledge on
involved mechanisms. Some well-evaluated mechanisms are not sufficiently addressed in current
treatment programs (e.g., the role of expectation). The role of emotion regulation, a concept
that has been found to be of relevance for all of these syndromes, also might not be sufficiently
addressed in current intervention proposals. After reviewing the neuroimaging literature, Landa
and colleagues (2012) concluded that psychotherapeutic and/or pharmacological interventions that
foster the development of affect regulation capacities will serve to modulate aberrantly activated
brain circuits and thus will be of particular benefit in the treatment of somatoform pain.
Psychological interventions assigned to help patients with medically explained versus medi-
cally unexplained somatic symptoms might show some overlap, but some differences also must be
considered. Frequently, patients in both groups initially have a more biomedical concept about
the cause of their complaints, and therapists need to offer examples of behavioral experiments
to demonstrate how psychological factors can influence the intensity and annoyance of somatic
complaints. However, if the somatic symptoms derive from a serious medical condition, this must
be also considered both in explaining illness models and in deriving intervention rationales with
patients. A patient suffering from diabetes and SSD needs a treatment plan that integrates not only
coping strategies for somatic symptoms, but also strategies for interpreting somatic sensations cor-
rectly (e.g., differentiating hypoglycemia from general dizziness), for adhering to medication, and
for reducing fears of developing threatening conditions (e.g., blindness). In contrast, a patient with
somatization syndrome might need interventions such as defocusing self-attention from bodily
perception, increasing awareness for environmental events and their influence on bodily well-
being (competition of cues), reducing physical avoidance behavior, and establishing a useful level
of physical activity. Although catastrophizing physical sensations might be fully inappropriate in
patients with somatic symptoms not caused by a biomedical condition, it is much more appropriate
in a patient who received cancer treatment just a few months ago and is still at increased risk for
recurrence of the disease. Thus, the role of specific treatment programs, techniques, and target
processes may vary considerably between the possible “phenotypes” of SSD.
CONCLUSIONS
The revision of somatoform disorders to somatic symptom and related disorders in DSM-5 ad-
dresses a serious need for reformulation. However, it leaves several problems unsolved, and it
must be considered a challenge for researchers specialized in this field. Therefore, we suggest
modifications and specifications that are based on DSM-5 criteria, that allow crossovers with past
360 Rief ·Martin

lines of research, and that are less subject to unclear assumptions. As long as the similarity between
mechanisms involved in medically explained and medically unexplained symptoms is not shown,
researchers should test this hypothesis, which implies continuing to use this distinction, at least as
a covariate. Pain research and research on hypochondriasis/illness anxiety disorder have revealed
very specific results; these lines of research merit continuation and should not be sacrificed to
DSM-5. The suggested modifications, which are based on DSM-5 and on empirically validated
research, offer parallel developments.
DISCLOSURE STATEMENT
The authors are not aware of any affiliations, memberships, funding, or financial holdings that
might be perceived as affecting the objectivity of this review.
LITERATURE CITED
Agar-Wilson M, Jackson T. 2012. Are emotion regulation skills related to adjustment among people with
chronic pain, independent of pain coping? Eur. J. Pain 16(1):105–14
Aggarwal VR, McFarlane GJ, Farragher TM, McBeth J. 2010. Risk factors for onset of chronic oro-facial
pain—results of the North Cheshire oro-facial pain prospective population study. Pain 149:354–59
Allen LA, Woolfolk RL. 2013. Somatization and conversion disorder. In A CBT Reference Guide, Vol. 2, ed.
W Rief. New York: Wiley. In press
Am. Psychiatr. Assoc. 2013. Diagnostic and Statistical Manual of Mental Disorders. Washington, DC: Am.
Psychiatr. Publ. 5th ed.
Bach M, Bach D. 1995. Predictive value of alexithymia: a prospective study in somatizing patients. Psychother.
Psychosom. 64:43–48
Bailer J, Witthöft M, Rist F. 2008. Psychological predictors of short- and medium-term outcome in individuals
with idiopathic environmental intolerance (IEI) and individuals with somatoform disorders. J. Toxicol.
Environ. Health Part A 71:766–75
Bantick SJ, Wise RG, Ploghaus A, Clare S, Smith SM, Tracey I. 2002. Imaging how attention modulates pain
in humans using functional MRI. Brain 125:310–19
Barrett B, Tyrer P, Tyrer H, Cooper S, Crawford MJ, Byford S. 2012. An examination of the factors that
influence costs in medical patients with health anxiety. J. Psychosom. Res. 73:59–62
Barsky AJ, Ahern DK. 2004. Cognitive behavior therapy for hypochondriasis. A randomized controlled trial.
J. Am. Med. Assoc. 291:1464–70
Barsky AJ, Orav J, Bates DW. 2005. Somatization increases medical utilization and costs independent of
psychiatric and medical morbidity. Arch. Gen. Psychiatry 62:903–10
Barsky AJ, Wool C, Barnett MC, Cleary PD. 1994. Histories of childhood trauma in adult hypochondriacal
patients. Am. J. Psychiatry 151:397–401
Barsky AJ, Wyshak GL. 1990. Hypochondriasis and somatosensory amplification. Br. J. Psychiatry 157:404–9
Barsky, AJ, Wyshak G, Klerman GL. 1990. The somatosensory amplification scale and its relationship to
hypochondriasis. J. Psychiatr. Res. 24:323–33
Bernstein MJ, Claypool HM. 2012. Social exclusion and pain sensitivity: Why exclusion sometimes hurts and
sometimes numbs. Personal. Soc. Psychol. Bull. 38:185–96
Bogaerts K, Janssens T, De Peuter S, Van Diest I, Van den Bergh O. 2010a. Negative affective pictures can
elicit physical symptoms in high habitual symptom reporters. Psychol. Health 25:685–98
Bogaerts K, Van Eylen L, Li W, Bresseleers J, Van Diest I, et al. 2010b. Distorted symptom perception in
patients with medically unexplained symptoms. J. Abnorm. Psychol. 119:226–34
Bogaerts K, Wan L, Diest IV, Stans L, Decramer M, Van den Bergh O. 2012. Peak-end memory bias in
laboratory-induced dyspnea: a comparison of patients with medically unexplained symptoms and healthy
controls. Psychosom. Med. 74:974–81

Bohn D, Bernardy K, Wolfe F, Hauser W. 2013. The association among childhood maltreatment, somatic
symptom intensity, depression, and somatoform dissociative symptoms in patients with ﬁbromyalgia
syndrome: a single-center cohort study. J. Trauma Dissociation 14:342–58
Bradford K, Shih W, Videlock EJ, Presson AP, Naliboff BD, et al. 2012. Association between early adverse
life events and irritable bowel syndrome. Clin. Gastroenterol. Hepatol. 10:385–90.e3
Brown RJ. 2004. Psychological mechanisms of medically unexplained symptoms: an integrative conceptual
model. Psychol. Bull. 130:793–812
Brown RJ, Brunt N, Poliakoff E, Lloyd DM. 2010. Illusory touch and tactile perception in somatoform
dissociators. J. Psychosom. Res. 69:241–48
Brown RJ, Skehan D, Chapman A, Perry EP, McKenzie KJ, et al. 2012. Physical symptom reporting is
associated with a tendency to experience somatosensory distortion. Psychosom. Med. 74:648–55
Browning M, Fletcher P, Sharpe M. 2011. Can neuroimaging help us to understand and classify somatoform
disorders? A systematic and critical review. Psychosom. Med. 73:173–84
Burton C, Weller D, Sharpe M. 2009. Functional somatic symptoms and psychological states: an electronic
diary study. Psychosom. Med. 71:77–78
Cella M, White PD, Sharpe M, Chalder T. 2013. Cognitions, behaviours and comorbid psychiatric diagnoses
in patients with chronic fatigue syndrome. Psychol. Med. 43:375–80
Chou R, Shekelle P. 2010. Will this patient develop persistent disabling low back pain? J. Am. Med. Assoc.
303:1295–302
Clark DM, Salkovskis PM, Hackman A, Wells A, Fennell M, et al. 1998. Two psychological treatments for
hypochondriasis. Br. J. Psychiatry 173:218–25
Cloitre M, Miranda R, Stovall-McClough KC, Han H. 2005. Beyond PTSD: emotion regulation and inter-
personal problems as predictors of functional impairment in survivors of childhood abuse. Behav. Ther.
36(2):119–24
Colloca L, Benedetti F. 2009. Placebo analgesia induced by social observational learning. Pain 144:28–34
Colloca L, Petrovic P, Wager TD, Ingvar M, Benedetti F. 2010. How the number of learning trials affects
placebo and nocebo responses. Pain 151:430–39
Colloca L, Sigaudo M, Benedetti F. 2008. The role of learning in nocebo and placebo effects. Pain 136:211–18
Connelly M, Keefe FJ, Afﬂeck G, Lumley MA, Anderson T, Waters S. 2007. Effects of day-to-day affect
regulation on the pain experience of patients with rheumatoid arthritis. Pain 131(1–2):162–70
Constantinou E, Bogaerts K, Van Diest I, Van den Bergh O. 2013. Inducing symptoms in high symptom
reporters via emotional pictures: the interactive effects of valence and arousal. J. Psychosom. Res. 74:191–
96
Creed F, Guthrie E, Fink P, Henningsen P, Rief W, et al. 2010. Is there a better term than “medically
unexplained symptoms”? J. Psychosom. Res. 68:5–8
Creed FH, Davies I, Jackson J, Littlewood A, Chew-Graham C, et al. 2012. The epidemiology of multiple
somatic symptoms. J. Psychosom. Res. 72:311–17
Crichton F, Dodd G, Schmid G, Gamble G, Petrie KJ. 2013. Can expectations produce symptoms from
infrasound associated with wind turbines? Health Psychol. In press
Davis KD, Moayedi M. 2013. Central mechanisms of pain revealed through functional and structural MRI.
J. Neuroimmune Pharmacol. 8:518–34
de Jong JR, Vlaeyen JW, van Eijsden M, Loo C, Onghena P. 2012. Reduction of pain-related fear and increased
function and participation in work-related upper extremity pain (WRUEP): effects of exposure in vivo.
Pain 153:2109–18
Dimsdale J, Sharma N, Sharpe M. 2011. What do physicians think of somatoform disorders? Psychosomatics
52:154–59
Dimsdale JE. 2011. Medically unexplained symptoms: a treacherous foundation for somatoform disorders?
Psychiatr. Clin. North Am. 34:511–13
Dimsdale JE, Creed F, Escobar J, Sharpe M, Wulsin L, et al. 2013. Somatic symptom disorder: an important
change in DSM. J. Psychosom. Res. 75:223–28
Enck P, Bingel U, Schedlowski M, Rief W. 2013. The placebo response in medicine: minimize, maximize or
personalize? Nat. Rev. Drug Discov. 12:191–204
362 Rief ·Martin

Somatic symptom disorder

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Viewers also liked

Viewers also liked (12)

Similar to Somatic symptom disorder

Similar to Somatic symptom disorder (20)

More from Paul Coelho, MD

More from Paul Coelho, MD (20)

Recently uploaded

Recently uploaded (20)

Somatic symptom disorder